Kragen Sitaker <kragen+pgsql@airwave.com> writes:
> On Mon, Jan 12, 2004 at 06:20:23PM -0500, Tom Lane wrote:
>> No; an OID collision would have occurred when you tried to create a
>> table. If two tables are present in pg_class then they have different
>> OIDs, and shouldn't have any conflicts in pg_statistic.
> How would that OID collision manifest? Do you think the error message
> might look similar?
Similar, but referring to pg_class_oid_index.
regards, tom lane